trhgquan / PS3E19

Kaggle PS3E19 - Forcasting Mini-Course Sales

PS3E19 - Forecasting Mini-Course Sales

Kaggle Playground Series - Season 3 Episode 19

Introduction

For this challenge, you will be predicting a full year worth of sales for various fictitious learning modules from different fictitious Kaggle-branded stores in different (real!) countries. This dataset is completely synthetic, but contains many effects you see in real-world data, e.g., weekend and holiday effect, seasonality, etc. You are given the task of predicting sales during for year 2022.

Final result: 52.8423 on private test set, ranking 923/1172

Result

Version	Features	Baseline Model	Result (Public)	Result (Private)
v1	country, store, product	Linear Regression	79.55318	80.74542
v2	country, store, product	XGBoost1	53.62858	55.38368
v3	country, store, product	XGBoost1	52.89421	54.62068
v4	country, store, product + observations with holiday	XGBoost	54.51476	52.8423

License

This project is licensed under The GNU GPL v3

Footnotes

1. Same features, same baseline model but different result due to random seed. ↩ ↩²